Smart way to verify the identity of the sound, based on neural network technique identification by voice

Abstract

Nowadays there are many researches being carried out on image and sound processing fields all around the world, which are mainly using artificial intelligence methods and different processing algorithms like DSP, genetic algorithm, neural network, etc. The objective of this article is creating an intelligent method for enhancing the capability of identification based on neural network technique. This method is based on training a proper network, and is capable of detecting and classifying different audio signals and finally learn the implications that the user introduces for each group of sounds in a limited scope. In this article, the network is trained by the audio signals of individuals and the objective of the network after the training is the separation of input signals and detection of the individual linked with each signal. Finally, the results were presented and their conformity was investigated.